Hi, I am Josias Hernandez, I am one of the BI interns here at InterWorks. Today, I would like to walk you through a Market Basket Analysis app I built in Sigma. Most retail dashboards focus on descriptive analytics, and they summarize what has already happened, but they don't tell us which products actually drive add on purchases or which customers are likely to buy next. That keeps us in a reactive mode instead of anticipating where future growth will come from. This app adds an intelligence layer on top of a standard sales and marketing dashboard by bringing Python based analytics directly into Sigma. So, let's jump into the app and see how these basket patterns help us find the next best customer to target. Here we are. Initially, you might have an ash bore like this. It's very basic with a filter, some KPIs, but we wanna add some Python code and some more in-depth analytics. I'm gonna filter my sample because usually these analysis are done with a lot of records. Before proceeding, we can configure the association rule mining, changing the main parameters, the confidence and the support we wanna get. And let's run the code. So, in the background, Python is running, and we are performing what is called association rule mining, which is a data mining technique that discovered if then patterns about items that frequently occur together. So, here we can see some examples. Let's see the meaning of this rule. This first rule is telling us that when customers buy coffee makers and classic jeans, about ninety eight percent of those baskets also include luxury handbags and sports sneakers. This pattern appears in roughly twenty seven percent of all baskets analyzed, and overall, the selected segment is three point six times more likely to buy this basket than typical shoppers. This way, we can decide what special offers and what promotions to send and to whom we want to send that. Let's find those clients. Once we get the name of the clients and the emails, we can generate emails using AI in their model languages. So here we can see the customers that bought coffee makers and classic jeans, but they haven't acquired luxury handbags and sports sneakers yet. We can generate the emails using AI, and once the emails are generated, we can decide either to approve the messages, to change the messages, or to send them straight to the customers. So, we will test each one of these features. I will approve the messages in Spanish, in English, and I want to change this one in Deutsch. I'm going to do something here. I'm going to paste the greeting four times so we can detect that message once we receive it. I'm gonna approve this one, and I'm gonna send them. You can send them individually, or you can decide to export the list to a CSV file to use this in your own email marketing tool, let's say Mailchimp or any other, Okay? Here you can see the body of the message. Our next step is to check the inbox. We can see here the five messages. The last one received is the message we edited with that long greeting. So, here is the message in French. Here are the messages in English. This and this one is in Spanish. Notice that the message is in HTML format. We changed that using another AI prompt that converted the messages from plain text to HTML. Now, we are ready to see our forecast. This improved version of the initial dashboard is including future information. We are assuming that the five emails were delivered and those customers clicked on email and purchased the articles we suggested to them. That's why we can display a conversion rate and a growth in the number of orders and the sales figures expected for those customers. I hope you enjoyed this walkthrough. We brought together Python code, AI generated content, and email features in Sigma to turn a basic self dashboard into a tool that helps us make decisions and not just look at numbers. Thank you very much for your time and attention, and I am happy to answer any questions or hear your feedback. Cheers!